# The Law Applicable to Secondary Liability in Intellectual Property Cases

**Authors:** Graeme B. Dinwoodie, Rochelle C. Dreyfuss, Annette Kur
**Citation:** "The Law Applicable to Secondary Liability in Intellectual Property Cases," 42 *N.Y.U. J. Int’l L. & Pol.* 201 (2009) (with Graeme B. Dinwoodie and Annette Kur)

**Abstract:** In recent years, intellectual property law has paid increasing attention to issues of private international law. The American Law Institute promulgated Intellectual Property: Principles Governing Jurisdiction, Choice of Law, and Judgments in Transnational Dispute in 2008. In Europe, the Max Planck Institutes’ Conflict of Laws in Intellectual Property Conflicts of Law effort is expected in 2010. However, neither of these projects has dealt explicitly with choice of law on contributory liability (or any other form of secondary liability that makes one party liable for the harm caused by another). Yet, actions premised on secondary liability are rapidly becoming the favored route for efficient enforcement on a worldwide basis. Examples include cases that attempt to impose liability on manufacturers of copying technologies for infringements caused by those who use their equipment; on purveyors of peer-to-peer file sharing software for the activities of those who download material without rightholders’ permissions; on internet service providers for subscribers’s infringing postings; and on other intermediaries, such as auction sites. In principle, secondary liability actions can occur in all areas of intellectual property law. However, for purposes of this paper, we concentrate on trademark cases, such as the litigation involving the responsibility of the online auction house, e-Bay, for the sale of counterfeit goods on its website. The problems posed in that area are particularly complex. After offering a stylized fact pattern to illustrate the problems, we consider the different ways in which courts might deal with questions arising in cases where secondary liability claims are asserted. We suggest that the traditional approach to choice of law in trademark cases generates unacceptable uncertainties for intermediaries and that a genuine engagement with conflicts scholarship would help mediate among the diverse interests and policy concerns. In the end, however, we conclude that private international law solutions may not resolve all the complications of multinational secondary liability cases. Thus, we are moved to propose, as an alternative solution, an autonomous (substantive) principle applicable in these cases. We conclude with some thoughts about how the different approaches engage with existing models for the resolution of trans-border intellectual property disputes and with the international intellectual property regime more generally.
**Source:** http://nyujilp.org/wp-content/uploads/2013/02/42.1-Dinwoodie-Dreyfuss-Kur.pdf

## I. INTRODUCTION

*p. 2*
In recent years, intellectual property law has paid increasing attention to issues of private international law. The problems involved in determining such issues as jurisdiction and choice of law have prompted at least two organizations to formulate general principles governing multi-state disputes. The American Law Institute (ALI) promulgated Intellectual Property: Principles Governing Jurisdiction, Choice of Law, and Judgments in Transnational Disputes in 2008. 1 In Europe, the Max Planck Institutes' Conflict of Laws in Intellectual Property (CLIP) effort is expected in 2011. 2 Although these projects deal with most private international law issues arising in international intellectual property litigation, neither has dealt explicitly with choice of law on contributory liability (or any other form of secondary liability that makes one party liable for the harm caused by another). 3 Yet, actions premised on secondary lia- ). The first preliminary draft is available at http://www.ip.mpg.de/shared/data/pdf/draft-clip-principles-08-04-2009.pdf. There are also parallel projects ongoing in Korea and Japan.

*p. 2*
3. Addressing questions of choice of law in cases of secondary liability runs headlong into what conflicts scholars would call the challenge of characterization (or, in some countries, "qualification"): when does a case involve "secondary liability"? This is an ever-present dilemma for private international law. But it is particularly acute in this context. Different countries use a variety of labels to denominate the different forms of secondary liability that exist. Indeed, courts in some countries do not characterize potential liability as "secondary," but rather talk of liability flowing directly from a failure to do certain acts (such as implement measures or engage in monitoring or supervision). See, e.g., LVMH v. eBay, Tribunal de commerce [T.C.P.] [court of trade] Paris, June 30, 2008, 11-12, (Fr.) available at http://web20. nixonpeabody.com/np20/np20wiki/PDF%20Library/Ebay1.pdf.; cf. LVMH v. Google, [cite] (Advocate-General Maduro, Sept. 22, 2009) (noting potential basis of intermediaries under national law). Moreover, even within countries, semantics can obscure understanding in this area. See Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417,435 (1984) (quoting Universal City Studios, Inc. v. Sony Corp. of America, 480 F. Supp. 429, 457-458 (C.D. Cal. 1979)) ("[T]he lines between direct infringement, contributory infringement, and vicarious liability are not clearly drawn."); Thomas C. bility are rapidly becoming the favored route for efficient enforcement. These actions permit rights holders to focus their attention on "deep pockets," save them the trouble of suing individual end-users (who may also happen to be their customers), and can provide a simple mechanism for obtaining worldwide injunctive relief. Examples include cases that attempt to impose liability on manufacturers of copying technologies for infringements caused by those who use their equipment, on purveyors of peer-to-peer file sharing software for the activities of those who download material without rights holders' permissions, on internet service providers for subscribers' infringing postings, and on other intermediaries, such as auction sites. 4 Folsom, Toward Non-Neutral Principles of Private Law: Designing Secondary Liability Rules for New Technological Uses, 3 AKRON INTELL. PROP. J. 45, 52 (2009) ("Secondary liability . . . might seem chaotic. Even the choice of words used to describe the genus, species and strands of the various related doctrines is not entirely free from doubt. Sometimes, 'secondary' liability refers to the genus, while 'vicarious' and 'contributory' refer to two of the species of secondary liability. This usage is by no means universal."). For example, in the United States, a fairly established doctrinal structure has been developed by courts to examine secondary liability in copyright infringement cases, with relatively clear (if contested) standards for contributory infringement, on the one hand, and vicarious infringement, on the other. But even there, it is unclear whether there is separate secondary liability for "active inducement," or whether that concept is simply a variant of the more well-established forms. See Jay Dratler, Palsgraf, Principles of Tort Law and the Persistent Need for Common Law Judgment, Perhaps Even on IP Infringement, 3 AKRON INTELL. PROP. J. (forthcoming 2009) (outlining the elements of contributory and vicarious infringement). In trademark law, the doctrines are even less clear, with far less judicial exploration of the differences between contributory and vicarious infringement. See Inwood Labs., Inc. v. Ives Labs., Inc., 456 U.S. 844, 853-54 (1982); Hard Rock Cafe Licensing Corp. v. Concession Servs., Inc., 955 F.2d 1143, 1148-52 (7th Cir. 1992). Thus, in this paper, we have used the term "secondary liability" as an umbrella term to encompass broadly the different bases under which an intermediary would be held liable not for unauthorized use of the plaintiff's mark, but for use of the mark by a third party. We do so fully cognizant of the difficulties of characterization that will ensue, but that is a problem intrinsic to this topic that will confront any proposed solution. 4. See, e.g., Metro-Goldwyn Mayer Studios, Inc. v. Grokster, Ltd., 545 U.S. 913 (2005) (claim against purveyors of peer-to-peer software); Sony Corp. v. Universal City Studios, 464 U.S. 417 (1984) (claim against manufacturers of video-cassette recorders); A&M Records, Inc. v. Napster, Inc., 239 F.3d 1004, 1010-1011 (9th Cir. 2001) (claims against distributors of peer-topeer software); Tiffany (NJ) Inc. v. eBay, Inc., 576 F. Supp. 2d [Vol. 42:201 These suits raise many of the same problems as appear in other transnational intellectual property litigation, where the territoriality principle makes it difficult to localize the entire dispute in a single jurisdiction and where the resulting proliferation of potential suits leads to forum shopping, uncertainty over liability, conflicting demands, and overlapping rights. However, secondary liability actions also create further complications in that the policies they implicate extend beyond those found in the context of direct infringement. Thus, secondary liability more directly involves the trade-off between "supporting creative pursuits . . . and promoting innovation in [the] new communication technologies [that make infringement possible]." 5 In principle, secondary liability actions can occur in all areas of intellectual property. However, they are most prevalent in cases involving digitally transmitted content-which is to say, trademarks and copyrighted material. 6 For purposes of this paper, we concentrate on trademark cases, such as the litigation involving the responsibility of the online auction house, eBay, for the sale of counterfeit goods on its website. 7 The problems posed in this area are particularly complex. Activity potentially giving rise to secondary liability may interfere with the signaling function of trademarks. At the same time, however, that activity allows sellers to reach larger markets and fa-(S.D.N.Y. 2008) (rejecting Tiffany's attempt to hold auction site liable for sale of counterfeit items on its website."); Cartoon Network LP v. CSC Holdings, Inc., 536 F.3d 121, 123 (2d Cir. 2008) (considering whether defendant cable company's digital video recorder system, allowing home recording of cable broadcasts, infringed plaintiff's copyrights). Indeed, some of the biggest battles in recent years have revolved around whether particular activity should be treated under principles of direct or secondary infringement. See, e.g., Cartoon Network LP v. CSC Holdings, Inc., 536 F.3d 121, 123 (2d Cir. 2008) (considering whether copying of copyrighted content was performed by cable company in ways that gave rise to direct liability); Stacey L. Dogan & Mark A. Lemley, Trademark and Consumer Search Costs on the Internet, 41 HOUS. L. REV. 777 (2004) (liability of search engines for sale of keywords under trademark law).

*p. 4*
5. Grokster, 545 U.S. at 928. 6. This is not to say that similar issues will not arise in the context of patent law. Patentees have recently begun to confront issues of digitized material crossing borders in ways that might adversely affect their rights. See, e.g., Microsoft Corp. v. AT&T, 550 U.S. 437 (2007); NTP, Inc. v. Research in Motion, Ltd., 418 F.3d 1282 (Fed. Cir. 2005).

*p. 4*
7. See, e.g., Tiffany, 576 F. Supp. 2d at 457. cilitates legitimate uses of marks by owners of legitimate products. These cases thus pit the marketing interests of trademark holders against the development of new mechanisms for trading in goods.

*p. 5*
After offering a stylized fact pattern to illustrate the problems, we consider the different ways in which courts might deal with questions arising in cases where secondary liability claims are asserted. We suggest that the traditional approach to choice of law in trademark cases generates unacceptable uncertainties for intermediaries and that a genuine engagement with conflicts scholarship would help mediate among the diverse interests and policy concerns. In the end, however, we conclude that private international law solutions may not resolve all the complications of multinational secondary liability cases. Thus, we are moved to propose, as an alternative solution, an autonomous (substantive) principle applicable in these cases. We conclude with some thoughts about how the different approaches engage with existing models for the resolution of trans-border intellectual property disputes and with the international intellectual property regime more generally. [Vol. 42:201 that simulate the appearance of a trademark holder's product which bear a spurious mark that is identical to the trademark holder's. 9 Sellers create their own description of the goods on the site. However, Z-Bay has an extensive monitoring program aimed at maintaining trust in its auction sites. These efforts include banning of sites that expressly offer counterfeit ("knock off") goods and a program called UnReAL, which allows trademark owners to report to Z-Bay any listing offering potentially infringing items, so that Z-Bay can remove items that are not genuine.

## II. TRADITIONAL INTELLECTUAL PROPERTY APPROACHES

*p. 6*
For trademark holders, Z-Bay presents at least two potential problems. First, the used goods placed on the market exert a downward pressure on the price at which new goods can be marketed. Second, counterfeit goods can disrupt the trademark holder's business and destroy its reputation and goodwill. These problems ultimately led two trademark holders to sue Z-Bay for injunctive and monetary relief: Epiphany sued in Xandia, Z-Bay's residence; 10 LuiV sued in Patria, where its own headquarters are located. In both cases, the plaintiff claimed that Z-Bay was contributorily liable for the infringements committed by those selling goods on its site.

*p. 6*
The Patrian court applied Patrian trademark law on direct and contributory liability and decided the case in favor of LuiV. Focusing on standard trademark law values, it held that an injunction was necessary to protect consumers against deception and producers against the loss of their investment in goodwill. Finding that Z-Bay should have done more to police its marketplace, it enjoined Z-Bay from permitting the sale of counterfeit goods as well as genuine goods first sold outside of Patria. The court also re- quired Z-Bay to ban used goods when they are advertised without a full disclaimer of connection with LuiV. In addition, it awarded damages based on past sales of these products worldwide.

*p. 7*
In contrast, the Xandian court applied Xandian trademark law on direct and contributory liability and decided the case in favor of Z-Bay. Focusing on issues of technological development and free trade, it held that an injunction would interfere with the development of new business models and would prevent consumers from freely alienating legitimate goods. Holding that Z-Bay had done enough to police its marketplace, the court also denied monetary relief. 11 Each of the decisions on liability involves a plausible interpretation of trademark law, and each court's approach to jurisdiction and applicable law reflects how courts tend to deal with transnational intellectual property disputes. Courts typically assume jurisdiction over claims asserted under their own trademark law; they ordinarily will not apply foreign trademark law or adjudicate foreign trademark rights. 12 And the fact that a decision may have some extraterritorial effect has not traditionally precluded courts from granting relief. 13 However, the interaction of these decisions is troubling from many perspectives. For Z-Bay, the problem is that it will be difficult to receive the benefits of its Xandian victory while complying with the Patrian judgment. It could segment the market and police each country's auction site according to that nation's law. 14 However that would destroy the benefits of aggregating buyers and sellers, and would restrict the sale of genuine trademarked goods. Alternatively, it could preserve the worldwide auction, but only by adopting the restrictive business model imposed by the Patrian court, leading to significant costs. 15 Most important, it faces significant uncertainties, for Epiphany, LuiV, or new trademark holders could sue in yet another jurisdiction, where the new court might impose even more burdensome monitoring requirements and award further monetary relief. 16 All these are also problems for customers: they could lose the benefits of the consolidated auction site or be forced to pay the costs of Z-Bay's enhanced enforcement obligations.

*p. 8*
From the perspective of the states involved, the outcome is not wholly satisfactory either, for each state can undermine the interests of the other. If Z-Bay adopts the restrictive model in order to comply with the Patrian decision, then Patria is essentially permitted to extrude its policies on a worldwide basis. Xandia's interest in supporting a global marketplace would be frustrated, as would be its domestic policy of fostering the development of so-called Web 2.0 technologies, which facilitate not only global marketing, but also such activities as global communication, collaboration, and social networking. It would also diminish innovation in related fields, such as de-14. Many global online commercial actors have adapted their business models to limit participation in their marketplace by consumers in one country or another because of, inter alia, different applicable legal norms. For example, some might refuse to send goods to a particular jurisdiction. However, compelling this type of behavior effectively requires the adoption of second-best solutions.

*p. 8*
15. See Hartford Fire Ins. Co. v. California, 509 U.S. 764, 799 (1993) (Souter, J.) (noting that a defendant can comply with two different judgments so long as one does not require action that the other prohibits).

*p. 8*
16. See, e.g., Microsoft Corp. v. Lindows.com, Inc., 319 F. Supp. 2d 1219 (W.D. Wash. 2004) (noting parallel actions in Europe on counterpart trademarks); Computer Assocs. Int'l, Inc. v. Altai, Inc., 126 F.3d 365 (2d Cir. 1997) veloping advertising models and payment schemes. 17 Furthermore, even if the law of Patria prevails, the decision may jeopardize Patria's long-term interests in that the outcome invites other countries to apply their own laws extraterritorially in later cases, without regard for Patria. 18 Of course, these two cases may lead Z-Bay to decide to segregate its auction sites. But in that case, the trading interests of both Xandia and Patria are damaged because their buyers and sellers lose the advantages of an abundant and diversified marketplace.

## III. ALTERNATIVE PROPOSALS

*p. 9*
In large part, all of these problems stem from a commitment to territoriality in the face of global commerce. The reluctance to adjudicate foreign trademark disputes means that parties are compelled to re-litigate national claims in a number of fora. And even if a single court can assume jurisdiction over a worldwide case, the territorial nature of intellectual property rights means that such an action would properly involve the assessment of a number of discrete national claims under numerous (and potentially different) national laws.

## A. Proposals by the ALI and CLIP

*p. 9*
Both the ALI and CLIP suggest that the solution to the territoriality problem is to make it easier to bring actions against direct infringers. First, they provide a greater opportunity for consolidation than presently exists. Under these proposals, it is likely that a trademark holder could bring all of its worldwide claims against a seller (or in some instances, sellers) in a single court. 19 [Vol. 42:201 tional cases of ubiquitous infringement, permit courts to derogate from the territorial approach and simplify the case by applying a single law to the entire dispute. For example, § 321 of the ALI Principles provides that: When the alleged infringing activity is ubiquitous and the laws of multiple States are pleaded, the court may choose to apply to the issues of existence, validity, duration, attributes, and infringement of intellectual property rights and remedies for their infringement, the law or laws of the State or States with close connections to the dispute. 20 These changes significantly improve the options for trademark owners. For example, under traditional trademark principles, if LuiV were to bring an action for direct infringement against the seller (or sellers) of counterfeit goods, it would be required to initiate a separate action in each country where an infringement occurred, alleging a violation of separate national trademark laws. In each court, LuiV would have to prove its ownership of a trademark and infringing acts by the seller. On both issues, courts typically apply their own domestic law, requiring LuiV to prove ownership of rights under the lex fori and (in many cases) local use of the LuiV mark by the seller in ways likely to cause confusion. However, despite early judicial suggestions, use online does not constitute use everywhere. 21 putes together in Xandia or Patria, and thus receive a single judgment binding upon both trademark holders. But this extended form of consolidation, involving numerous disparate parties, was not at the core of the consolidation provisions in the ALI Principles and the diversity of parties might make the argument for consolidation harder to sustain, especially in countries where the notion of broad multiparty, multistate litigation is less common. For the analogous provisions of the CLIP Principles, see CLIP, supra note 2, R arts. 2:102 (general jurisdiction); 2:202 (special jurisdiction on infringement); 2:203 (extent of jurisdiction); 2:205 (multiple defendants).

*p. 10*
20. See ALI PRINCIPLES, supra note 1, § 321(1). Article 3:603(1) of the CLIP PRINCIPLES, supra note 2, similarly provides that:

*p. 10*
In disputes concerned with infringement carried out through ubiquitous media such as the Internet, the court may apply the law or the laws of the State or the States having the closest connection with the infringement, if the infringement arguably takes place in every State in which the signals can be received. 21. Although the question of where a seller "used" a mark when acting online initially presented difficulties for many national courts-for example, regarding whether the seller "uses" the mark only where it uploads the ad- Each national court would be able to grant an injunction and award damages only in respect of the national claim before it based on the use in that country.

*p. 11*
In contrast, the ALI and CLIP projects would allow LuiV to save litigation costs by consolidating all of its separate national claims against direct infringers in a single action. Some hope (and others might fear) that the adjudication by a single court of similar claims under multiple national laws might exert a de facto centripetal force on applicable law and that a court hearing such a case would apply one law (normally its own) to the entire dispute. 22 However, instead of relying on subconscious judicial assimilation, both the ALI Principles and the CLIP Project take a proactive role. Because the case involves ubiquitous infringement, they would permit LuiV to simplify the case and apply a single law to the worldwide dispute. 23 vertisement, or where the advertisement could be viewed-a growing consensus suggests that the question of where use occurs must be answered with an eye to where commercial effects are felt. That is the solution adopted by 23. The relevant provision of the CLIP Principles is not as conducive to application in trademark cases because it is limited to situations where "the infringement arguably takes place in every State in which the signals can be received." CLIP art. 3:603. However, it would likely apply in cases of well known marks, where rights are essentially ubiquitous. [Vol. 42:201 This approach recognizes that, despite historical commitments to territoriality, the application of numerous national laws to conduct that bears no overriding connection to a single state is not only inefficient but arguably also relatively weak in prescriptive legitimacy. 24 To identify the single applicable law, it is necessary to consider multiple connecting factors. Section 321 of the ALI Principles suggests that, in determining the applicable law(s) in cases of ubiquitous infringement, courts should look at:

*p. 12*
(b) where the parties' relationship, if any, is centered;

*p. 12*
(c) the extent of the activities and the investment of the parties; and (d) the principal markets toward which the parties directed their activities. 25 In an action against a single seller, it may appear that these factors point equally strongly to the residence of the seller and of the trademark holder. However, this is not the case. On Z-Bay, the seller is directing its activity to the entire world and the buyers are not likely to know where the seller resides. Accordingly, the strongest connections are with Patria, the residence of LuiV, the trademark holder. 26 The con- 26. The CLIP Principles may lean more strongly than the ALI Principles towards applying the law of the place where the alleged infringer is located, arguably casting some doubt on the invariable application of the law of Patria. But even under the CLIP Principles, Patria remains the most likely applicable law. Moreover, the applicable law may, to some extent, be shaped by the architecture of the cause of action. See Richard Garnett & Megan Richardson, Libel Tourism or Just Redress? Reconciling the (English) Right to Reputation with the (American) Right to Free Speech in Cross-Border Libel Cases, 5 J. PRIV. INT'L LAW (2009) (forthcoming) (discussing Gutnick), which is largely within the control of the plaintiff trademark owner. Thus, a trademark owner seeking to diminish the claims of the law of the place of a single seller might (in clusion in favor of Patria is even more obvious in actions against multiple sellers, who potentially reside all over the world. However, even under these innovative approaches, there are problems. Both the ALI Principles and the CLIP Principles contain safety features designed to reflect the continued territorial nature of intellectual property laws. For example, the ALI would permit the court to apply the laws of more than one country, if those laws have sufficient connection with the dispute. In that case, the court would apply each of the laws chosen to the relevant part of the case. 27 In addition, the ALI regime would allow the parties to carve out specific jurisdictions and demonstrate that the law of that particular state differs from the law chosen. In such cases, relief would be fashioned to deal, for example, with states where the trademark is not recognized or which have very different rules on exhaustion. 28 These provisions improve the accuracy of the outcome and give greater deference to the interests of states other than the one with the closest connection to the case. But they do make adjudicating the case more complicated and interfere with the possibility of arriving at a global solution.

*p. 13*
An even more significant problem is that-especially given the complexities involved in suing direct infringerstrademark holders may find the simplifications provided by these projects insufficient. Accordingly, they are still likely to prefer the efficiencies of suing a single defendant-namely, Z-Bay-under a theory of secondary liability. 29 If LuiV were to addition to emphasizing the ubiquity of the buyers) try to consolidate sellers in other jurisdictions.

*p. 13*
27. ALI PRINCIPLES, supra note 1, § 321(1). 28. ALI PRINCIPLES, supra note 1, § 321(2); cf. Playboy Enters., Inc. v. Chuckleberry Corp., 939 F. Supp. 1032, 1040 (S.D.N.Y. 1996) (injunction in the United States did not bar defendant publisher from maintaining its Internet site in other countries where plaintiff did not possess trademark rights, but could be invoked to limit access of Americans to web site hosted in Italy). CLIP also permits the parties to prove that particular state laws differ from the one chosen by the court. sue Z-Bay for conduct that enabled infringement occurring in a multitude of countries, the case would once again be consolidated (for jurisdictional reasons, this time, most likely in Xandia). However, regardless of where the dispute was consolidated, the prevailing assumption of intellectual property scholars has been that the law governing the claim of contributory infringement should follow that applicable to the alleged acts of direct infringement. 30 Neither the ALI Principles nor the CLIP project challenges this assumption. Under this approach, then, the claim would again be regarded as relating to ubiquitous infringement, leading the court to apply § 321.

*p. 14*
Since contributory infringement depends on the law applicable to direct infringement, the Xandian court would apply Patrian law, for the reasons set out above. Indeed, it is even clearer here that the law of Patria will apply because now both sellers and buyers are located in many different countries, diluting the significance of the law of any single seller's residence. 31 Fusing the law applicable to direct and secondary liability is problematic for a number of reasons. Viewed through the lens of contacts between the plaintiff and the direct infringer, 30. See ALI PRINCIPLES, supra note 1, § 301 cmt. h ("In some States, facilitation (or 'authorization') of infringement may itself be an infringing act . . . [i]n such cases, the law that governs the determination of primary infringement applies."); cf. Folsom, supra note 3 (exploring the nexus between a R primary question of direct liability and a secondary question of whether a third party should be held responsible for the acts of the direct infringer); Graeme W. Austin, Importing Kazaa-Exporting Grokster, 22 SANTA CLARA COM-PUTER & HIGH TECH. L.J. 577, 596 (2006) (discussing root copy theory and suggesting that "[t]hough the lines between various forms of domestic infringement may be blurry . . . geopolitical lines are not: the territoriality principle should preclude application of U.S. liability theories where the primary acts of infringement occur abroad.").

*p. 14*
31. Z-Bay might in theory avoid the broadest application of Patrian law by availing itself of the safeguards provided by the ALI and CLIP principles.

*p. 14*
See supra text accompanying notes 27-28 (discussing safeguards). Thus, it R could argue that while its conduct is considered to be contributory infringement under Patrian law, Xandian law imposes less demanding standards in that respect, thereby forcing the court to take Xandian law into account in fashioning the remedy. However, such a carve-out will hardly lead to meaningful results for Z-Bay, because, in practice, it will mean that Z-Bay must limit its auction site to Xandia (or any other country or countries where it can be proven that the legal situation corresponds to that in Xandia), thereby disrupting its specific business model. the conflicts analysis may be incomplete. 32 For direct infringement, the parties' interests will ordinarily serve as proxies for state interests. For example, the interest of the trademark holder will capture the state's interest in encouraging investment in goodwill. However the additional state interests implicated more directly by contributory (or any other form of secondary) infringement may not be fully captured by looking simply at the parties involved in the direct infringement. For instance, an interest in regulating emerging technologies is, in reality, an interest in fostering the entry of new market players, who will often be competitors to the parties before the court.

*p. 15*
Taking these other interests into account favors the application of Xandian law to the case against Z-Bay.

*p. 15*
To be sure, one solution to an incomplete list of connecting factors would be to supplement § 321. For example, the list of connecting factors could be expanded to include the interests of the states where the relevant technology was being developed or where the intermediary was based. However, adding new connecting factors to accommodate additional concerns relevant to secondary liability might have the effect of diluting factors that should have more determinative weight in the context of direct infringement.

*p. 15*
To put this another way, the strategy of enlarging the list of connecting factors to accommodate the possibility of secondary infringement claims merely highlights the basic problem in the approach taken by both projects: because the projects conceptualize secondary liability as derivative of primary liabil-32. Moreover, the claim against the direct infringer might be characterized as ubiquitous less easily than might the contributory claims against Z-Bay. Likewise, depending on the standard for contributory infringement (in particular, whether a court would look at whether the defendant's service facilitated infringement only in the case of the trademark before it, or more systemically with respect to trademarked products more generally), choosing a single law applicable to the entire dispute might be difficult. Even if a single law is identified, a party unhappy with the law chosen might seek refuge within Section 321(2), which provides that "notwithstanding the State or States designated pursuant to subsection (1), a party may prove that, with respect to particular States covered by the action, the solution provided by any of those States' laws differs from that obtained under the law(s) chosen to apply to the case as a whole. The court shall take into account such differences in determining the scope of liability and remedies." ALI PRINCIPLES, supra note 1, § 321(2). ity ("facilitating infringement" 33 ), they make the law on one issue inappropriately hostage to the law on the other. This intellectual-property-centric view (arguably grounded in legitimate intellectual property policies) fails to consider non-traditional intellectual property interests or whether, as in other tort contexts, the two types of liability might be determined under different laws. 34 B. An Autonomous Tort Approach: Determining Law Applicable to Secondary Liability

## Grounds For and Operation of a Tort-Based Conflicts Rule

*p. 16*
A simpler-albeit more radical-solution to the problems raised by the hypothetical Z-Bay case would be to recognize that contributory infringement actions are, at bottom, torts and treat them autonomously, rather than as supplementary forms of intellectual property infringement actions. Like the result under the ubiquitous infringement provisions of the ALI and CLIP projects, a single law will apply to the claim against Z-Bay. However, as the choice of law applicable will not automatically follow the law applicable to direct infringement, courts could embark on a more nuanced analysis. The choice would properly reflect all of the policy concerns at play in contributory infringement cases. For example, the court would now include in the analysis the interest in facilitating worldwide sales, developing new business models, and encouraging technological innovation. This approach conceptualizes contributory infringement claim as a tort that serves different social purposes from the tort actions that underlie direct infringement liability. 35 In fact, this approach is consistent with the way the laws of secondary infringement appear to be developing. For example, the justification for secondary copyright infringement is arguably different from the rationale for direct infringement. 36 The copyright statute-unlike the patent statute-makes no separate, express provision for secondary liability. However, in Grokster, the U.S. Supreme Court partially grounded the development of contributory and vicarious copyright infringement in common law principles that extend the scope of potential liability to persons not engaged in acts of primary infringement. 37 And the policy reasons for sustaining-or resistingsuch expansions go beyond the considerations typically at play in determining primary infringement. 38 Similarly, the Australian courts view "authorization" as a discrete cause of action (a 35. In addition, stated in the language of contemporary conflicts analysis, secondary liability rules might be characterized as loss-allocating as opposed to conduct-regulating. See Tkaczevski, 22 F. Supp. 2d at 173. Questions of direct infringement centrally involve the interests of the state where conduct occurs. See Itar-Tass Russian News Agency v. Russian Courier, 153 F.3d 82 (2d Cir. 1998) (holding that issues involving alleged infringement of copyrights in works created by Russian nationals and first published in Russia would be governed by U.S. law where the location of alleged infringement was the United States); cf. Vanity Fair Mills, Inc. v. T. Eaton Co., 234 F.2d 633, 639 (2d Cir. 1956) (noting that "the wrong takes place where the passing off occurs, i.e., where the deceived customer buys the defendant's product in the belief that he is buying the plaintiff's"). Loss-allocating rules do not as directly implicate the place of conduct. view also held by some courts in the United States). 39 While these courts have made the substantive move to consider secondary liability as a separate tort, they have not explicitly addressed the choice of law question. 40 Under what we are calling "the autonomous tort approach," LuiV could sue Z-Bay under a single law that is more appropriately attuned to the concerns at play in determining liability for contributory infringement. In a jurisdiction that employed interest analysis, the relevant factors might simply mirror those enumerated in § 321. However, the party relevant to the analysis would be Z-Bay and the trademark holders, rather than the direct infringers. It would also be possible to refine the analysis further. Thus, the factors could be expanded so that courts choosing the law applicable to secondary liability would consider:

*p. 18*
(b) where the parties' relationship, if any, is centered;

*p. 18*
(c) the extent of the activities and the investment of the parties; 2). In the United States, the majority approach rejects an autonomous cause of action for authorization of infringing acts. See Subafilms, Ltd. v. MGM-Pathe Comms. Co., 24 F.3d 1088, 1094 (9th Cir. 1994) (en banc). But this approach is not without critics. See Curb v. MCA, 898 F. Supp. 586, 595 (M.D. Tenn. 1995).

## A recent decision by Stockholm Tingsrätt [TR] [Stockholm District

*p. 18*
Court] 2009-04-17 ref B 13301-06 (Swed.) adopts a similarly unengaged approach to copyright infringement, albeit with some recognition of cross-border implications. The case primarily involved the criminal liability of persons hosting a large bit-torrent tracker. Id. at 16. However, the judgment also addressed civil liability of the defendants because a number of rights holders-partly based in Scandinavia, partly in the United States-had claimed and obtained substantial damages. Id. at 15-16. The court applied Swedish law (without discussing the issue of applicable law), id. at 46, but deducted a certain percentage of the damages claimed to have been caused by illegal downloading, because, as the decision put it, those acts "also occurred outside of Sweden." Id. at 70-71. (f) the nature of the primary activity alleged to be infringing; and (f) the place or places where the technological innovations at issue were created. 41 In countries adhering to the lex loci delicti rule to determine the law applicable to tort actions, identifying the locus of the separate tort of contributory infringement would likely lead to the same result as under interest analysis informed by an expanded list of connecting factors. 42 There are several other ramifications of adopting an autonomous tort approach to secondary liability. First, this approach truly conceives of the problem as global and provides a global solution determined by a single law. There is no need to resort to the complications of determining whether infringement is ubiquitous, whether-despite ubiquity-more than one law applies, or whether there is a need to carve out specific jurisdictions from the scope of relief. Second, separating the causes of action for analytical purposes makes clear that rights holders can pursue a secondary infringement claim without necessarily bringing the suit in a jurisdiction where direct infringers can be joined. 43 Third, intermediaries will find their conduct judged by a law that is likely open to their concerns about technological development and new business models. Fourth, while Z-Bay might be vulnerable to multiple 41. These factors in ALI Section 321 are similar to those enumerated in the RESTATEMENT (SECOND) OF CONFLICTS OF LAWS § 145 (1971), which is the general rule for determining applicable law in tort actions. See id. at § 145(2) (listing "the place where the injury occurred, the place where the conduct causing the injury occurred, the domicile, residence, nationality, place of incorporation and place of business of the parties, and the place where the relationship, if any, between the parties is centered" as "contacts to be taken into account" in determining "the law applicable to an issue" under Section 145 lawsuits and no new trademark holder would be estopped from re-litigating the issues, its vulnerability to further suits is likely reduced for reasons of comity or stare decisis. At the very least, the same trademark holder could not bring serial actions against Z-Bay in a number of different countries, hoping to find a hospitable forum. 44 Fifth, because the liability of intermediaries will likely be more limited, trademark holders now have more reason to pursue their own actions against direct infringers. In these actions, they are likely to be litigating under a set of rules more favorable to their interests and to trademark values generally. 45 Thus, in these direct infringement suits trademark values will be of paramount consideration not only in determining the applicable law, but also in deciding the case.

*p. 20*
To put this another way, one problem with the result if the Patrian court simply viewed the question of secondary liability through the lens of the actions for direct infringement is that it could easily ignore the benefit of a viable secondary market in trademarked goods. When the case is broken into two parts, this does not happen. Where the focus is on the intermediary, all of the activity on the website tends to be considered as a whole. At the same time, however, genuine trademark interests are not ignored. Indeed, they are more carefully considered: in the action against direct infringers, the court can engage in a more fine-grained analysis and distinguish between counterfeiters and legitimate resellers. 46 The end result is that the market for the sale of used goods is less likely to be suppressed.

*p. 20*
Notwithstanding these systemic effects, courts (and some intellectual property scholars) have insisted that a finding of direct infringement is a necessary predicate to making out a case for secondary liability. Some of these concerns seem formalistic, insisting that logic dictates a particular order for try-44. See Microsoft Corp. v. Lindows.com, Inc., 319 F. Supp. 2d 1219, 1221 (W.D. Wash. 2004) (noting that Microsoft had pursued litigation in seven states).

*p. 20*
45. The tort move might also change the jurisdictional analysis because the contracts relevant for determining specific jurisdiction may now be different.

*p. 20*
46. Courts could further distinguish among counterfeits, grey goods, and goods in respect of which the trademark holder's rights have been exhausted. ing and deciding a case. 47 But experience suggests that these arguments do implicate real policy concerns. For example, using equitable trust doctrines, the U.S. courts have used the existence of a "root copy" or "predicate act" in the United States to assume prescriptive authority over allegedly infringing activity abroad that flows from the U.S. conduct. 48 Courts applying this theory have not typically begun their analysis by asking for proof that the acts abroad were infringing under the foreign law in question. 49 Likewise, the Digital Millennium Copyright Act prohibits the sale of circumvention tools regardless of whether they are being used to enable specific acts of copyright infringement. 50 Because the root copy cases have thus resulted in copyright holders securing damages for acts abroad 47. See Subafilms, Ltd. v. MGM-Pathe Communications Co., 24 F.3d 1088 (9th Cir. 1994); Columbia Pictures Indus. v. Prof'l Real Estate Investors, Inc., 866 F.2d 278 (9th Cir. 1989). 48. See, e.g., L.A. News Serv. v. Reuters Television Int'l Ltd., 149 F.3d 987, 991-92 (9th Cir. 1998) (unauthorized transmission and copy of work made in the United States and then further transmitted to Europe and Africa), later proceedings, L.A. News Serv. v. Reuters Television Int'l Ltd., 340 F.3d 926, 931-32 (9th Cir. 2003); Update Art, Inc. v. Modiin Publ'g, Ltd., 843 F.2d 67, 72-73 (2d Cir. 1988) (unauthorized copy of plaintiff's poster made in the United States and then further copied and distributed in Israel); Sheldon v. Metro-Goldwyn Pictures Corp., 106 F.2d 45, 52 (2d Cir. 1939) (awarding plaintiff profits from both U.S. and Canadian exhibition of infringing motion picture where a copy of the motion picture had been made in the United States and then shipped to Canada for exhibition), aff 'd, 309 U.S. 390 (1940). 49. See, e.g, Sheldon, 106 F.2d at 52. But cf. Filmvideo Releasing Corp. v. Hastings, 668 F.2d 91 (2d Cir. 1981) (declining to enjoin foreign exhibition absent proof that foreign copyright laws were infringed). 50. See, e.g., Universal City Studios, Inc. v. Corley, 273 F.3d 429, 443 (2d Cir. 2001). More recently, courts have deployed a number of interpretations of the statute to defeat efforts to use the DMCA to repress competition in the market for uncopyrightable products and services, and thus have insisted (at a general level) on some nexus between the rights protected by copyright law and the technological protection measures that are alleged to have been unlawfully circumvented. See, e.g., Chamberlain Group, Inc. v. Skylink Techs., Inc., 381 F.3d 1178, 1192-93 (Fed. Cir. 2004); Lexmark Int'l, Inc. v. Static Control Components, Inc., 387 F.3d 522, 545 (6th Cir. 2004); Storage Tech. Corp. v. Custom Hardware Eng'g & Consulting, Inc., 421 F.3d 1307, 1318 (Fed. Cir. 2005) [Vol. 42:201 not proved to be infringing, they extrude U.S. national copyright law to the global market. 51 The DMCA similarly tends to expand the scope of rights holders' control over possibly noninfringing uses. 52 However, as noted at the outset, secondary liability for intellectual property infringement is, in many ways, more than just another means to ensure efficient enforcement of rights. Rather, the regulation of the behavior of intermediaries can conceivably be regarded as a new objective of intellectual property law, raising autonomous concerns. Certainly, the possibility of acts of infringement that span national borders has intensified. And some legal regimes have created supplementary statutory causes of action to prevent national borders from becoming a tool for undermining intellectual property rights. 53 The autonomous tort approach to conflicts is also analytically quite different from the theory underlying root copy liability. The real complaint about the root copy cases is that they fail to take into account the potentially different interests of the states where the primary activities occurred. For example, courts in these cases rarely consider whether the subsequent distributions are infringing under the laws of the countries of distribution. 54

## Remaining Problems With an Autonomous Tort Approach

*p. 23*
Admittedly, this approach does not solve all of the problems at play in the Z-Bay case. Even if it allows courts to choose a conflicts rule that will lead to the application of a single law applicable to the contributory infringement analysis, 56 Z-Bay's interests are not fully protected. Because the core problem for Z-Bay is its inability to efficiently monitor those who are using its website, it remains vulnerable to suits by many other trademark holders, none of whom are bound by the original judgment. Although the adoption of a uniform conflicts rule would theoretically lead to the same result in each case, Z-Bay will face some continuing uncertainty due to the malleability of the conflicts analysis. There are two procedural ways to resolve this problem. One is to provide that suits may not be brought in states other than the one that resolved the initial dispute. (In the Z-Bay case, that would likely be Z-Bay's country of domicile.) The other approach is to allow Z-Bay to aggregate actions against all potential plaintiffs for declarations of non-liability.

*p. 23*
The first solution is feasible only if there is general agreement on a personal jurisdiction rule that restricts the defendant's amenability to suit. It is unlikely that most countries would renounce the use of domestic courts to protect their theory); Graeme W. Austin, The Infringement of Foreign Intellectual Property Rights, 113 LAW Q. REV. 321 (1997) (analyzing the basis for jurisdictional prohibition of actionability of claim of infringement of foreign intellectual property rights).

*p. 23*
55. Moreover, because this approach involves the articulation of a choice of law rule, the lex causae might still insist on there having to be a showing of direct infringement. Indeed, under any conceivable rule regarding secondary liability, it is unlikely that the standard for secondary infringement could be satisfied without some primary infringement having occurred. The more pertinent question is whether there are sufficiently substantial noninfringing uses being facilitated by the intermediary that we should ignore the primary infringements that are also facilitated.

*p. 23*
56. This approach leads to the adoption of a single law using traditional conflicts principles, such as lex loci delicti and interest analysis, without explicit reliance upon the relatively untested notion of ubiquitous infringement. Thus, it also demonstrates that the ubiquitous infringement approach may not be as radical as some believe. citizens from local claims. The second solution is equally unrealistic. As with the first solution, it would require some trademark holders to submit to a jurisdiction with which they have little relationship (after all, the trademark holders did not put their own goods up for auction on Z-Bay, and thus they do not have contacts with its residence). Additionally, such a procedure is essentially an action against a class of defendants. U.S. law contemplates defendant class actions. 57 However, they are rarely utilized and it is highly doubtful that such a procedure would ever be recognized or adopted internationally.

## C. A Substantive Approach to Secondary Liability

*p. 24*
A strength of the autonomous tort approach is that, as a conflicts rule, it gives broad scope to national interests. A weakness, however, is that this gives countries latitude to impose their parochial interests above those of the interests of an international marketplace. Some countries (Patria) may overvalue trademark holders' interests and sacrifice international interests in developing new technologies that integrate world commerce. Others may be so supportive of intermediaries' business interests that they offer insufficient protection to trademark rights. 58 A uniform substantive approach to secondary liability might therefore be preferable.

*p. 24*
It is not unknown in intellectual property to adopt substantive international norms in order to avoid fragmentation through the application of private international law. In some cases, these international instruments limit national prescriptive authority by articulating a substantive rule that displaces any inconsistent national provision. For example, the EC Satellite and Cable Directive defines the act of communication to the public as occurring only in the country in which material is 57. FED. R. CIV. P. 23(a) (permitting a class to "sue or be sued"). 58. The TRIPs Agreement by its terms leaves open the question of WTO members' obligations to protect against contributory infringement. The enforcement chapter of the Agreement requires members to ensure that "enforcement procedures as specified [in the Agreement] are available under their law so as to permit effective action against any act of infringement of intellectual property rights . . . including expeditious remedies to prevent infringements and remedies which constitute a deterrent to further infringements." TRIPs Agreement, art. 41(1). But it does not dictate that such measures include actions for secondary infringement. uploaded for transmission to the public. The Directive thus avoids the consequences of overlapping power in every country in the footprint of the transmitter. 59 Soft law can also act in this manner, promulgating a norm that becomes the governing substantive international standard without formally overriding national legislative prerogatives. For example, the World Intellectual Property Organization (WIPO) Joint Recommendation Concerning Provisions on the Protection of Marks, and Other Industrial Property Rights in Signs on the Internet suggests limiting prescriptive authority to the states in which the use of a trademark on the Internet has a "commercial effect." 60 In addition, the Recommendation protects online actors from overbroad remedial obligations by cautioning national courts not to enjoin behavior on a worldwide basis unless it is proportionate to the harm committed. 61 Furthermore, soft law approaches have been adopted to facilitate effective worldwide enforcement of intellectual property rights when particular conduct is seen as clearly violative of the international trademark norms. For example, the Uniform Dispute Resolution Policy of the Internet Corporation for Assigned Names and Numbers (ICANN's UDRP) sets out a series of factors for determining when a party has engaged in In some respects, the rule in the Cable and Satellite Directive can be conceptualized as a harmonized choice of law rule, dictating in which state the transborder act will be deemed to occur. 60. Joint Recommendation Concerning Provisions on the Protection of Marks, and Other Industrial Property Rights in Signs, on the Internet, art. 2, WIPO Doc. 845(E) (Oct. 2001) [hereinafter Joint Recommendation on Internet Use] (adopted by Assembly of the Paris Union for the Protection of Industrial Property and General Assembly of the World Intellectual Property Organization), available at http://www.wipo.int/about-ip/en/development_ iplaw/doc/pub845.doc. The Recommendation defines commercial effect in art. 3. In a similar vein, the CLIP draft principles state that an infringement shall only be found where the defendant has substantially acted in furtherance of the infringement or where the infringement has substantial effect. See CLIP, supra note 2, art. 3:602 ("De Minimis Rule").

*p. 25*
61. See Joint Recommendation on Internet Use, supra note 60, at art. 15 (urging states to avoid remedies that would prohibit all use on the internet); see also id., Explanatory Note (discussing remedies). cybersquatting. 62 The soft law character of the UDRP and its choice of law provision both emphasize the continued importance of national adjudication and national law. 63 However, in effect, the UDRP has served as the de facto substantive rule governing cybersquatting disputes in many countries. 64 In part, this is because WIPO's members agreed that a certain subset of uses of marks clearly violate the international norm against unfair competition found in article 10 bis of the Paris Convention. 65 As a result, there is now a subset of clearly infringing behavior that is effectively governed by an international norm. In contrast, online disputes between competing legitimate national trademark owners have remained a matter of national law. 66 These principles-proportionality, and targeting behavior widely agreed to violate international norms-form the framework for a potential substantive rule on secondary liability. All member states of the World Trade Organization have agreed . 24, 1999), available at http://www. icann.org/en/udrp/udrp-rules-24oct99.htm (authorizing panels to decide cases using "any rules and principles of law that [they] deem applicable"); UDRP, supra note 62, at ¶ 4(k).

*p. 26*
64. See Council for Trade-Related Aspects of Intellectual Property Rights, Electronic Commerce Work Programme: Submission from Australia, IP/C/W/233, ¶ 44 (Dec. 7, 2000), available at http://docsonline.wto.org/DDFDocuments/ t/IP/C/W233.doc (suggesting that the UDRP has "arguably become a de facto international standard" for the resolution of cybersquatting disputes).

*p. 26*
65. Final Report of the WIPO Internet Domain Name Process, ¶ 173-74 (Apr. 30, 1999) (noting that WIPO's proposed definition of abusive registration "draws on solid foundations in international and national law and in case law," and citing in particular unfair competition provisions of the Paris Convention).

*p. 26*
66. See Department of Commerce, Management of Internet Names and Addresses, 63 Fed. Reg. 31,741, 31,747 (June 10, 1998) to prohibit the sale of counterfeit products. 67 In contrast, the conditions under which trademark rights are exhausted by the sale of a legitimate trademarked item vary from one country to another. For example, some countries follow a regime of national or regional exhaustion; other countries adhere to a theory of international exhaustion. The TRIPS Agreement, the highest statement of international intellectual property norms, expressly refused to take a position on this point. 68 Countries are thus free to follow either approach. Moreover, countries nominally adopting the same approach to the geographic scope of exhaustion can take a different approach to the resale of goods that have been altered in some way or that are presented to the public in a form that suggests endorsement or association with the trademark owner. 69 Following the approach of targeting behavior widely regarded as actionable at the international level, while leaving more contested questions to national law, suggests that an international rule for intermediaries' liability might hold that:

*p. 27*
An [Vol. 42:201 This rule would offer trademark holders an efficient enforcement mechanism against counterfeiting not currently guaranteed by international law. They can sue Z-Bay in cases where it is actively inducing counterfeit sales (subsection (a)). They can also provide Z-Bay with notice of counterfeiting and force Z-Bay to remove counterfeit goods. If new technically and commercially feasible monitoring technologies are developed by trademark holders in the future, courts could require intermediaries to implement those measures to avoid the imputation of constructive knowledge (subsection (b)). Furthermore, rights holders can enforce any court orders obtained against direct infringers (subsection (c)). At the same time, the rule affords considerable protection for Z-Bay in that it would only be liable for facilitating activity that clearly violates international norms. 72 Further, this solves the logical problem noted in connection with the conflicts rules in that it premises secondary infringement liability on counterfeiting, and hence on direct infringement.

*p. 28*
A substantive international approach would not, however, solve all of the problems for Z-Bay. Had the Patrian and Xandian courts operated under this rule, they could have still adjudicated the cases differently because they could have come to differing conclusions on what level of knowledge is required and what sorts of monitoring measures are feasible. However, it is to be hoped that over time, as decisions become accessible throughout the world and as judicial networks evolve, outcomes will converge. 73 For example, national courts have in large part acted consistently with the Joint Recommendation on Internet Use on crafting relief in online cases. 74 This approach would also leave some problems for rights holders. They would lose the option of using secondary infringement as a mechanism for obtaining worldwide relief for forms of infringement other than counterfeiting. For example, LuisV and Epiphany could not sue Z-Bay for facilitating the sale of parallel imports. Trademark holders are not, how- 1103, 1124 (2000) (discussing the rise of networks of judges). 74. See, e.g., Playboy Enters. v. Chuckleberry Publ'g, Inc., 939 F. Supp. 1032, 1036-37 (S.D.N.Y. 1996). ever, without their remedies. In the short term, they can pursue actions against direct infringers. To the extent that these infringements will likely qualify as ubiquitous, a small set of laws will likely apply, making the case easier to litigate. Since the interests in developing new technology would not be at issue, the classic trademark policy issues would be salient in these actions. Because subsection (c) is not restricted to counterfeiting, a trademark holder could also force Z-Bay to remove from its website any goods that a court had found to be infringing under national law, such as goods that the jurisdiction regarded as not subject to exhaustion. Complementary measures could be taken to mitigate the remaining hardships for rights holders. For example, suits against direct infringers would be more efficacious if the intermediaries engaged in secondary infringement were required to provide information about the use and users of their systems. 75

## IV. SITUATING THE PROPOSALS IN THE INTERNATIONAL INTELLECTUAL PROPERTY REGIME

*p. 29*
The three approaches discussed in the previous section each contain innovations in the treatment of intermediary liability for trademark infringement in the global marketplace. Each challenges the broader premises of the international intellectual property regime. Some echo other features of the international system; others build and improve upon existing practices.

## A. ALI and CLIP Proposals

*p. 29*
At present, most national courts will only permit consolidation of copyright claims; trademark claims and-even more so-patent claims must be litigated on a country-by-country basis. In all cases (including copyright), the resolution of multi-state disputes depends on the application of multiple domestic laws. The ALI and CLIP proposals liberalize both aspects of the current system, permitting consolidation of all kinds of intellectual property cases and allowing the application of a single law in the case of ubiquitous infringement. These ideas are hardly radical. The consolidation proposals build on recent developments in copyright law worldwide and European efforts to provide a more efficient system for enforcing patent rights. 76 For example, a decade ago, the Second Circuit approved the adjudication of eighteen different national claims in a worldwide copyright licensing dispute. 77 Likewise, the application of a single law to aspects of a worldwide dispute, although hardly the norm, is not unknown. In some instances, the application is (as previously noted) subconscious. Indeed, the French court in the LVMH v. eBay case, which inspired the hypothetical, applied French law to worldwide claims, albeit without considering the international ramifications of its holding. In other cases, courts readily localize international activity in the forum. Internet cases are particularly susceptible to this analysis because it is relatively easy to conceptualize access as local conduct. 78 However, in some cases the approach represents implicit recognition that the international marketplace requires the application of a single law to particular issues. For example, the Second Circuit has effectively adopted something close to a lex originis approach to the question of initial ownership of copyrights (and thus applied a single law to the question of ownership worldwide). 79

## B. The Autonomous Tort Approach to Choice of Law

*p. 31*
The debate over choice of law in online intellectual property disputes is characterized by concerns over "information havens" and "information hells." The concern is that a country with lax protection will use the inevitable happenstance of effects within its borders to enable worldwide unauthorized distribution of protected material, or that a state with overbroad rules will take advantage of the same phenomenon to restrict uses that other states would regard as permissible. The ALI and CLIP largely address this problem through rules on personal jurisdiction. The autonomous tort approach to choice of law provides a new lever in the context of intermediaries. Trademark owners can sue direct infringers under laws congenial to their interests. On the other hand, by situating themselves in countries that are hospitable to technological and business innovation, intermediaries enhance the probability that their conduct will be assessed under rules more favorable to them. This enables them to offer an efficient mechanism for global trade. The result "splits the baby": neither the information haven nor the information hell can dominate the world market. The approach thus vindicates concerns for national autonomy. Regulatory diversity also promotes another form of technological innovation: by allowing states to recognize claims that require intermediaries to adopt novel enforcement tools, regulatory diversity encourages innovations in compliance mechanisms. Both the ideas of national autonomy and the role of nations as laboratories are components of the classical intellectual property system.

*p. 31*
The autonomous tort approach advances the private international law of intellectual property by engaging with contemporary conflicts scholarship. In particular, its regard for the different substantive purposes of primary and secondary infringement law takes into account the lessons of interest 79. Itar-Tass Russian News Agency v. Russian Kurier, Inc., 153 F.3d 82 (2d Cir. 1998). The ALI adopts a similar approach, especially in copyright. See ALI Principles, § 313. See, e.g., Steele v. Bulova Watch Co., 344 U.S. 280, 285-86 (1952); Vanity Fair Mills, Inc. v. T. Eaton Co., 234 F.2d 633 (2d Cir. 1956). In recent cases, some courts have stressed the importance of viewing such questions with an eye to the substantive purposes of trademark law. See McBee v. Delica Co. 417 F.3d 107, 111 (1st Cir. 2005) (holding that in order for a plaintiff to reach foreign activities of foreign defendants in American courts, the "complained-of activities [must] have a substantial effect on United States commerce, viewed in light of the purposes of the Lanham Act."). See generally Dinwoodie, supra note 54. These tests, which more closely resemble the approach of § § 402-403 of the Third Restatement of Foreign Relations Law than the Second Restatement of Conflicts, have not, however, given rise to particularly nuanced analysis by trademark courts. 81. 153 F.3d 82 (2d Cir. 1998). The Second Circuit's reliance on the Second Restatement of Conflicts, id. at 90, to develop a copyright choice of law rule as a matter of federal law inevitably injected some contemporary conflicts material into its statement of the rule it adopted. The court thus referred to an " 'interest' approach," id. at 91, and the law of the state with "the most significant relationship" to the property and the parties, id. at 90. However, the application of those concepts remained wooden, unlike the more nuanced regard for underlying state interests found in (the better) judicial explorations of the Second Restatement.

*p. 32*
82. See, e.g., Rome II Regulation, supra note 42 (largely endorsing the lex loci protectionis). For a discussion of the problems with strict application of lex loci protectionis, see MIREILLE VAN PROBLEMS (1965). For a list of the factors relevant to the choice of applicable rule of law, see RESTATEMENT (SECOND) OF CONFLICTS OF LAWS § 6 (1971). See also RESTATEMENT (THIRD) OF THE FOREIGN RELATIONS LAW OF THE UNITED STATES § § 402-403 (1987) (noting factors relevant to the determination of an unreasonable exercise of jurisdiction to prescribe law). While the two prior approaches, which are based on private international law, ameliorate some of the problems of territoriality in a global marketplace, they do not ensure an optimal scope of liability for intermediaries that operate worldwide. Diverse national laws will continue to determine the values according to which the conduct of intermediaries will be judged. 84 In contrast, articulating a substantive international approach will erase the potential for intermediaries to be governed by divergent norms, thereby creating a global standard. The global standard proposed above reconnects primary and secondary liability, which is regarded by many as a logical imperative. It also gives technological innovators room to experiment, gives rights holders incentives to invent compliance mechanisms, and enables the development of efficient global markets without disregarding the core objectives of national trademark rights. As such, it is fully consistent with the stated purposes of the TRIPS Agreement.

*p. 33*
Another way to conceptualize the substantive approach is as setting a ceiling on the scope of protection permitted under the TRIPS Agreement. Because TRIPS is a minimum-rights regime and-like the previous international conventions-includes very few mandatory limitations, 85 it is largely regarded as a one-way ratchet. It permits nations to continually expand their own intellectual property rights and to extract commitments from other countries to do likewise in bilateral trade agreements. For example, in a Free Trade Agreement with Australia, the United States has succeeded in destroying the flexibilities TRIPS provides on the question of exhaustion. 86 84. Moreover, the ability of Z-Bay to ensure the application of Xandian law to its activities under our autonomous tort approach might be affected by whether a Patrian trademark owner is able to frame a claim in a way that concentrates the contacts with its place of residence, allowing a Patrian (or another) court room to find Patrian law applicable even under our proposed choice of law rule.

*p. 33*
85. See, e.g., Berne Convention for The Protection of Literary and Artistic Works, Sept. 9, 1886, as last revised at Paris on July 24, 1971, 1161 U.N.T.S. 3.

*p. 33*
86. United States-Australia Free Trade Agreement, U.S.-Austl., art. 17, May 18, 2004, available at http://www.ustr.gov/sites/default/files/uploads/ agreements/fta/australia/asset_upload_file469_5141.pdf. In a way, that agreement aims at doing some of what trademark holders are seeking in the eBay cases: limiting the scope of the secondary market in trademark goods. The substantive approach, which limits secondary liability for the resale of legitimate goods, is a counterweight to these efforts. Discussions of international ceilings are ongoing in a number of fora; 87 this proposal advances these efforts in a nuanced way. In exchange for limiting the exposure of intermediaries to secondary liability, it strengthens the mechanism for enforcing agreed norms against counterfeiting. It imposes a global norm with respect to global intermediaries, but reserves some area of local autonomy for dealing with direct infringers.

*p. 34*
V. CONCLUSION Fifteen years ago, the TRIPS Agreement put intellectual property front and center in the development of global trade. No sooner was the ink dry, than it became clear that a genuine global market in knowledge products would be heavily dependent on digital communication technologies-information conduits, methodologies for exchanging content (be it scholarly, entertaining, or political), venues that bring geographically diverse parties together to share common interests, services that facilitate commercial transactions. Protecting these internet service providers (such as AOL and universities), peer to peer file sharing facilities (such as SSRN), web hosting services (TWEN), search engines (Google), data aggregators (Yahoo! News), chat rooms (lawProf), social networking cites (Facebook), and auction sites (eBay) has become a precondition to the rich and vibrant "knowledge economy" in which we now operate.

*p. 34*
States typically will have diverse views on how to adapt long-held values to the new environment. Xandia may want to foster the dynamism of this marketplace by giving latitude to these online entrepreneurs. Yet, Patria could have a more static vision, one that sees a more prominent role for traditional intellectual property interests. Under traditional principles, mediating these differences presents apparently intractable dilemmas for trans-border intellectual property litigation. COPYRIGHT 35-49 (2008), http://www.ivir.nl/publications/hugenholtz/limitations_ex-ceptions_copyright.pdf (recommending a global instrument on limitations and exceptions in the international copyright system). However, the silver lining might be that it has forced intellectual property law to confront the complexity of private international law.

*p. 35*
The paper has explored the strengths and weaknesses of private international law as a means for solving the problems posed by the global online marketplace. It has also demonstrated that even with a well-developed private international law, some matters might benefit from the adoption of substantive international norms. As Andy Lowenfeld has long insisted, a full understanding of the international system needs to take account of both. It must temper strong calls to sovereignty (what intellectual property law calls territoriality) to achieve the benefits of a truly global society.

## Footnotes

> 36. See Lipton, supra note 17, at 14; see also Mark Bartholomew, Copyright, R Trademark and Secondary Liability After Grokster, COLUM. J. L. & ARTS. (forthcoming 2009) (Buffalo Legal Studies Research Paper Series, Paper No. 2009-06 at 20-21, available at http://ssrn.com/abstract=1329058); Folsom, supra note 3, at 54, 59-60. R 37.